Research On Similarity Of Weibo Users’ Attention Bias

With the rapid development of online social networking platforms,the number of users keeps rising.The platforms represented by Sina Weibo have grown into influential industry leader.Weibo has been fully integrated into people’s daily life with its features of simple and easy operation,fast information dissemination and strong autonomy.It is of great significance to study the attention bias of Sina Weibo users for the platform to improve service quality and user satisfaction.Considering the differences in the degree of attention bias of Weibo users and the diversity of attention events,this paper proposes a similarity research method of user attention bias and conducts experiments in the actual network of Weibo.The main research contents and results are as follows:1.User similarity calculation method combining user attention relationship and user behavior.This paper argues that both user relationship and user behavior reflect the attention bias among users.There are similarities in the attention bias of users with attention relationships,and the similarity of attention bias among users with active behavior is higher than that of users with inactive behavior.Combined with the existing research,the user relation is used to calculate the similarity of concern relation,the standardized Euclidean distance method is used to calculate the similarity of activity,and the final user similarity is obtained by combining the two similarity.2.User attention is biased to mining.This paper selects representative users in Weibo as the benchmark users.Calculates the similarity between benchmark users,and uses the similarity combined with Louvain algorithm to mine the community structure of benchmark users to obtain benchmark communities.Calculate the similarity between the target user and benchmark community,and then study the attention bias of the target user.3.Use real Weibo data for verification.According to the experimental results,the method in this paper can find the similarity of the attention bias between the target user and the benchmark user.This similarity is shown by the user similarity index.Its size can show the similarity degree of attention between target users and benchmark community,and the diversity of attention of target users,so that users’ attention bias can be more comprehensively displayed.The experimental results show that this method can mine the similarity of attention bias among users,and get the degree of attention bias between users and benchmark user groups.The degree of attention bias can be measured and compared by user similarity.